Deciphering the correlation between Chr3q26.2 amplification through homologous recombination deficiency and improved survival in ovarian cancer under platinum-based adjuvant therapy.

Abstract

e17537 Background: Discriminant copy number alterations (CNAs) have been linked to aberrant DNA repair mechanisms. This study aimed to identify CNA-based biomarkers, alongside homologous recombination deficiency (HRD) scores for ovarian cancer patients receiving platinum (Pt)-based chemotherapy, especially in BRCA1/2 wild-type or HRD-low patients. Methods: Fourteen most prevalent and significant focal CNA regions (seven amplification and seven deletion) were examined to investigate their correlations with HRD scores via SHAP values of an XGBoost model and the effectiveness of Pt-based adjuvant chemotherapy in a cohort of high-grade serous ovarian cancer (HGSC) patients (TCGA cohort, N=576), and the results were validated in a separate Chinese ovarian cancer cohort (test cohort, N=457). An in-house developed next-generation sequencing HRD pipeline (GeneseeqPrime HRD) was used to determine HRD scores. Histological subtypes, molecular features, and potential molecular mechanisms associated with the identified biomarkers were further explored. Results: Of 14 high-level CNA regions identified by GISTIC in the TCGA cohort, Chromosome 3q26.2 amplification [ Chr3(q26.2) Amp, a prevalence of 30.7%] was associated with increased HRD scores in both BRCA1/2-altered ( p=0.01) and wild-type HGSCs ( p=0.03). Consistently, in the test cohort, higher HRD scores ( p<0.01) and elevated chromosome instability scores ( p<0.01) were observed in Chr3(q26.2) amplified samples. Among 406 patients administrated with Pt-based adjuvant chemotherapy in the TCGA cohort, those with Chr3(q26.2) Amp exhibited a relatively high sensitivity rate ( p=0.09) and significantly superior disease-free survival after adjusting for HRD score, patient age, tumor stage, and residual disease [hazard ratio (HR): 0.65, 95% confidence interval (CI): 0.48–0.88] when compared to those without, representing Chr3(q26.2) Amp as an independent predictive biomarker. Similar results were obtained in 78 patients using Pt-based adjuvant chemotherapy in the test cohort (HR: 0.68, 95% CI: 0.40–1.15). Moreover, Chr3(q26.2) Amp was more prevalent in HGSCs harboring BRCA1/2 or TP53 aberrations than other histological subtypes driven by PIK3CA, PTEN or KRAS mutations ( p<0.01). In comparison to HGSCs without Chr3(q26.2) Amp, Chr3(q26.2) amplified HGSCs exhibited higher Ragnum hypoxia scores ( p<0.01) and a total of four overexpressed proteins, including p62 ( p<0.01), PIK3CA ( p<0.01), CCNB1 ( p<0.01) and TSC1 ( p<0.01), indicative of suppressed homologous recombination repair pathway and cycle cell arrest. Conclusions: Chr3(q26.2) Amp emerges as a promising biomarker for Pt-based adjuvant chemotherapy in ovarian cancer, potentially reflecting a hypoxia-induced etiology through suppressed homologous recombination repair pathway.